N.B. The email exchange below may have been edited, e.g. to remove content not essential to the main point(s) or to standardize English spelling/grammar.
...I can only wonder why no lexicon sorts the Indo-European terms by meaning.
Doing this is difficult, in that it may be hard for a reader to guess or predict what English word(s) may have been selected to represent meanings; despite this, we DO "sort the Indo-European terms by meaning." On our page Indo-European Lexicon: Pokorny Master PIE Etyma, click the "Semantic Index" link on the left, and drill down two levels... The 1st set of "(3) (2) (1)" links on a line [2 levels down] leads to an etymon in a Pokorny Master PIE table; any 2nd set of "(3) (2) (1)" links on such a [semantic category] line leads to a collection of IE reflex words derived from said etymon.
The words used in our index terms are from Carl Darling Buck's A Dictionary of Selected Synonyms in the Principal Indo-European Languages (U. Chicago Press, 1949). But whereas Buck indexed reflex words by their [evolved] meanings, we index PIE roots directly and allow their reflexes to fall, rightly or wrongly, into the same semantic field/category as the root.
P.S. (July 21) As implied above, Buck sorted Indo-European terms by their meanings; in his case these terms were reflex words, not the PIE etyma themselves. In the Encyclopedia of Indo-European Culture by J.P. Mallory and D.Q. Adams, eds. (1997), PIE etyma are discussed in articles that deal with their meanings -- said articles being listed alphabetically according to the English meaning terms Mallory and Adams select, e.g. GRASS, MAKE, or STIFF.